RESEARCH ARTICLES
D. King, J. Roughgarden, Theor. Popul. Biol. 21, 194 (1982).
D. Cohen, J. Theor. Biol. 12, 119 (1966).
D. King, J. Roughgarden, Theor. Popul. Biol. 22, 1 (1982).
P. Haccou, Y. Iwasa, Theor. Popul. Biol. 47, 212 (1995).
E. Kussell, S. Leibler, Science 309, 2075 (2005).
L. Watson, M. J. Dallwitz, The Families of Flowering
Plants: Descriptions, Illustrations, Identification,
and Information Retrieval. Version: 29 July 2006
http://delta-intkey.com (1992 onward).
34. T. Keller, J. Abbott, T. Moritz, P. Doerner, Plant Cell 18,
598 (2006).
35. A. C. Whibley et al., Science 313, 963 (2006).
36. S. Gavrilets, Fitness Landscapes and the Origin of
Species, S. Levin, H. Horn, Eds., Monographs in
Population Biology (Princeton Univ. Press, Princeton and
Oxford, 2004).
37. We thank J. Avondo for help with visualizing 3D
fitness landscapes, M. Clauss for helpful discussions
on bet-hedging, and C. Thébaud for advice on inflorescence
databases. This research was funded by grants from Human
Frontier Science Program (E.C. and P.P.), Natural Sciences
and Engineering Research Council of Canada (P.P. and
L.H.), and the Biotechnology and Biological Sciences
Research Council, UK (E.C.).
Marine Radiocarbon Evidence
for the Mechanism of Deglacial
Atmospheric CO2 Rise
Thomas M. Marchitto,1,2*† Scott J. Lehman,1,2* Joseph D. Ortiz,3
Jacqueline Flückiger,2‡ Alexander van Geen4
We reconstructed the radiocarbon activity of intermediate waters in the eastern North Pacific over
the past 38,000 years. Radiocarbon activity paralleled that of the atmosphere, except during
deglaciation, when intermediate-water values fell by more than 300 per mil. Such a large decrease
requires a deglacial injection of very old waters from a deep-ocean carbon reservoir that was
previously well isolated from the atmosphere. The timing of intermediate-water radiocarbon
depletion closely matches that of atmospheric carbon dioxide rise and effectively traces the
redistribution of carbon from the deep ocean to the atmosphere during deglaciation.
adiocarbon measurements of calendrically dated hermatypic corals (1) and
planktonic foraminifera (2, 3) indicate
that the radiocarbon activity (D14C) of the atmosphere during the latter part of the last glacial
period [~20,000 to 40,000 years before the present (yr B.P.)] ranged from ~300 to 800 per mil
(‰) higher than it was during the prenuclear
modern era (Fig. 1C). Although reconstructions
of Earth’s geomagnetic-field intensity predict
higher cosmogenic 14C production rates during
the glacial period, production was apparently not
high enough to explain the observed atmospheric
enrichment (2–5). Rather, a substantial fraction of
the atmosphere’s D14C buildup must have been
due to decreased uptake of 14C by the deep
ocean. This requires a concomitant 14C depletion
in a deep-ocean dissolved inorganic C reservoir
that was relatively well isolated from the
atmosphere. Renewed ventilation of this reservoir could theoretically explain the drop in
atmospheric D14C (Fig. 1C) and the rise in atmospheric CO2 (6) across the last deglaciation.
Most workers point to the Southern Ocean as a
R
1
Department of Geological Sciences, University of Colorado,
Boulder, CO 80309, USA. 2Institute of Arctic and Alpine
Research, University of Colorado, Boulder, CO 80309, USA.
3
Department of Geology, Kent State University, Kent, OH
44242, USA. 4Lamont-Doherty Earth Observatory, Columbia
University, Palisades, NY 10964, USA.
*These authors contributed equally to this work.
†To whom correspondence should be addressed. E-mail:
[email protected]
‡Present address: Environmental Physics, Institute of Biochemistry and Pollutant Dynamics, Eidgenössische Technische
Hochschule Zürich, 8092 Zürich, Switzerland.
1456
locus of deglacial CO2 release, based on the similarity between atmospheric CO2 and Antarctic
temperature records (6) and on numerous conceptual and numerical models (7–9). If correct, we
would expect some signature of the low-14C deepocean C reservoir to be spread to other basins via
Antarctic Intermediate Water (AAIW). Here, we
report a strong radiocarbon signal of the deglacial
release of old C, recorded in an intermediatedepth sediment core from the northern edge of
the eastern tropical North Pacific.
Intermediate water D14C reconstruction.
Marine sediment multi-core/gravity-core/pistoncore triplet from sediment layer MV99-MC19/
GC31/PC08 was raised from a water depth of
705 m on the open margin off the western coast
of southern Baja California (23.5°N, 111.6°W)
(10). The site is today situated within the regional
O2 minimum zone that exists because of a
combination of high export production and poor
intermediate-water ventilation. Various sediment
properties in MC19/GC31/PC08 vary in concert
with the so-called Dansgaard-Oeschger (D-O)
cycles that characterized the Northern Hemisphere climate during the last glacial period (11).
Originally discovered in Greenland ice cores,
D-O cycles also exist in a number of lower-latitude
locations that were probably teleconnected to the
North Atlantic region through the atmosphere
(2, 12, 13). Off the coast of Baja California, the
sedimentary concentrations of organic C, Cd, Mo,
and benthic foraminifera all decreased sharply
during D-O stadials (cold periods in Greenland)
(11, 14). Together, these proxies are consistent
with reduced productivity during stadials, caused
8 JUNE 2007
VOL 316
SCIENCE
Supporting Online Material
www.sciencemag.org/cgi/content/full/1140429/DC1
Materials and Methods
SOM Text
Fig. S1
References
SOM Programs
25 January 2007; accepted 30 April 2007
Published online 24 May 2007;
10.1126/science.1140429
Include this information when citing this paper.
by either decreased coastal upwelling or a
deepening of the regional nutricline related to the
mean state of the tropical Pacific (11).
Diffuse spectral reflectance (DSR) provides a
1-cm resolution stratigraphy for GC31/PC08.
After R-mode factor analysis, the third factor of
DSR (Fig. 1A) exhibits the strongest correlation
to the productivity proxies and to Greenland climate (11). We used this DSR record to apply a
calendar-age model to MC19/GC31/PC08, based
on correlation to d18O (an air-temperature proxy)
in Greenland ice core GISP2 (Greenland Ice
Sheet Project 2) (15). Resulting calendar ages
were then combined with 50 benthic foraminiferal radiocarbon ages [19 of which were published previously (10)] to calculate age-corrected
intermediate-water D14C (16). To evaluate the
partitioning of 14C between the atmosphere and
the ocean, we compared intermediate-water D14C
to that of the atmosphere (Fig. 1C), as reconstructed from tree rings (17), U-Th–dated corals
(1, 17), and planktonic foraminifera from Cariaco
Basin off Venezuela (3). Calendar ages for
Cariaco Basin were originally based on the correlation of lithologic climate proxies to the
GISP2 d18O record (2), which has been layercounted with visual and chemical techniques
(15). However, Hughen et al. (3) recently
demonstrated that the Cariaco Basin 14C calibration yields much better agreement with coral
results older than ~22,000 yr B.P. when an
alternate age model is used, based on correlation
to the U-Th–dated Hulu Cave speleothem d18O
record from eastern China (13). Because DSR in
GC31/PC08 is more similar to the Greenland
isotope record than to the lower-resolution Hulu
Cave record, we continued to use the GISP2
correlation but applied simple provisional age
adjustments to GISP2 older than 23,400 yr B.P.,
using four tie points to Hulu Cave (Fig. 1B and
fig. S1). We do not suggest that this age model is
necessarily superior to the original one (15), but
this exercise is necessary for comparing our data
to the most recent (and most consistent) atmospheric D14C reconstructions (1, 3, 17). The
resulting age model for MC19/GC31/PC08,
based on 21 tie points, yields a very constant
sedimentation rate (fig. S2) and gives us
confidence that our calendar-age assignments
for 14C samples between tie points are reliable
to within a few hundred years (table S1).
Baja California intermediate-water radiocarbon activities are plotted in red in Fig. 1C. The
www.sciencemag.org
Downloaded from www.sciencemag.org on June 7, 2007
28.
29.
30.
31.
32.
33.
modern activity, based on a local seawater measurement of –131‰ at 445 m and the nearest
Geochemical Ocean Section Study profile (18),
is estimated to be –170‰, in good agreement
with the core top value. Comparable offsets
from the contemporaneous atmosphere were
maintained throughout the Holocene (typically
~100‰ between 0 and 10,000 yr B.P.) and
during the latter part of the glacial period
(roughly 200‰ between 20,000 and 30,000 yr
B.P.) (19). Radiocarbon activities before 30,000
yr B.P. show increased scatter that may be
related to the greater influence of slight contaminations on older samples.
Overall, it is clear that intermediate waters
mainly followed atmospheric D14C over the past
40,000 yr, except during the last deglaciation.
Activities dropped sharply just after 18,000 yr
B.P., reaching minimum values of ~–180‰
between 15,700 and 14,600 yr B.P., roughly
450‰ lower than that of the contemporaneous
atmosphere. For comparison, the lowest D14C
values found in the modern ocean (in the North
Pacific near 2-km water depth) are depleted by
~240‰, relative to the preindustrial atmosphere
(18). A second comparably large depletion
event began sometime between 13,500 and
12,900 yr B.P. and ended between 12,100 and
11,600 yr B.P. The magnitude of 14C depletion
we reconstructed during these two deglacial
events is much too great to attribute to changes
in dynamics of the North Pacific thermocline
(20) and must instead record a large change in
the initial 14C activity of waters advected to the
site. Such depleted waters could have been
sourced only from the deepest, most isolated
regions of the glacial ocean (21).
Link to atmospheric history. In Fig. 2, we
compare the timing of reconstructed intermediatewater and atmospheric D14C changes with the
deglacial record of atmospheric CO2 from the
East Antarctic Dome C ice core (6). The latter is
shown on a GISP2 layer-counted time scale
based on a simple synchronization of CH4
variations between the Dome C and GISP2 ice
cores (supporting online material text and table
S2). It is immediately apparent that the atmospheric D14C decline and CO2 rise occurred in
parallel, with a synchronous, intervening plateau
appearing in both records. There is a slight (and
still unreconciled) difference in the timing of the
start of the deglacial atmospheric D14C decline
between the Cariaco Basin (3) and IntCal04 (17)
reconstructions, but we take the deglacial onset
of the atmospheric CO2 rise and the atmospheric
14
C decline to be essentially synchronous. We
now also find that these changes were associated
with a prominent decline in D14C of intermediate
water in the eastern North Pacific that must
record the redistribution of aged C from the
deep ocean to the surface. After 14,600 yr B.P.,
intermediate-water activities rebounded to higher
values, coincident with the plateau in both the
atmospheric CO2 rise and the atmospheric D14C
drop. The leveling of the atmospheric records
and the increase in 14C activity of intermediate
waters are all indicative of a reduction in the flux
of aged C to the upper ocean and atmosphere
from below. After ~12,800 yr B.P., the atmospheric CO2 rise and D14C drop resumed, and
intermediate-water activities again reached minimum values of ~–180‰, indicating a resumption of C redistribution from the deep ocean to
the surface. By 11,500 yr B.P., the large deglacial
atmospheric shifts were mostly completed, and
intermediate-water activities finally reached modern values.
As pointed out by Monnin et al. (6), the
deglacial rise of atmospheric CO2 closely followed the rise in East Antarctic temperatures
(Fig. 3A), implying that the ocean’s release of C
to the atmosphere was associated with changes in
the Southern Ocean. Deep convection of the
Southern Ocean both ventilates much of the
ocean interior and returns to the atmosphere
much of the C extracted by photosynthesis in the
sunlit surface of the global ocean. During the last
Fig. 1. Intermediate-water and atmospheric D14C records. (A) Diffuse spectral
reflectance factor 3 from Baja California
composite sediment core MV99-GC31/
PC08, plotted versus depth (top axis)
(11). Gray lines show tie points to the
Greenland record that were used to derive
the calendar-age model. (B) d18O of
Greenland ice core GISP2 (15) on a
provisional revised time scale (black,
bottom axis). The new time scale deviates
from the original time scale (green) for
calendar ages older than 23,400 yr B.P.
(vertical line labeled G). The new time
scale is based on linear interpolation
between point G and three tie points
whose ages are derived from U-Th–dated
Hulu Cave (vertical lines labeled H) (13).
(C) Atmospheric radiocarbon activities
based on tree rings, planktonic foraminifera from Cariaco Basin varve-counted
sediments, and U-Th–dated corals (dark
green) (17); additional recent coral measurements (light blue) (1); and planktonic
foraminifera from Cariaco Basin, based on
an age model derived from the correlation
of sediment reflectance to Hulu Cave
(dark blue) (3). Red circles show
intermediate-water activities from benthic
foraminifera in MC19/GC31/PC08. The
yellow circle is an estimate for modern
bottom waters at this site. Error bars are
based on compounded uncertainties in
radiocarbon ages and calendar ages.
www.sciencemag.org
SCIENCE
VOL 316
8 JUNE 2007
Downloaded from www.sciencemag.org on June 7, 2007
RESEARCH ARTICLES
1457
glacial period, density stratification of the Southern Ocean surface and/or extensive sea-ice
coverage are suggested to have isolated deep
waters from the atmosphere (7–9), permitting the
buildup of a larger deep-ocean C reservoir and a
consequent drawdown of atmospheric CO2.
Sediment pore water chlorinity and d18O measurements, combined with benthic foraminiferal
d18O, indicate that deep Southern Ocean waters
were the saltiest and densest waters in the glacial
ocean (22). Such high salinities point to brine
formation beneath sea ice as an important mode of
formation. At deglaciation, a progressive renewal
of deep convection or upwelling in association
with documented sea-ice retreat (23) [and possibly
with poleward-shifting westerlies (8)] would have
provided for the simultaneous delivery of ocean
heat and sequestered C to the atmosphere. This
transition occurred in two major steps, beginning
with relatively early (~18,000 years ago) and
gradual increases in temperature and CO2 that
were temporarily interrupted by the Antarctic Cold
Reversal (ACR). Major transients in our record of
D14C in intermediate-depth waters of the eastern
North Pacific conform to this Antarctic schedule,
consistent with the redistribution of C from the
abyss to the upper ocean and atmosphere in
connection with changes in deep convection of
the Southern Ocean.
d13C provides a tracer that is complementary
to D14C, though with a far smaller dynamic
range in seawater. During the last glacial period,
the deep Southern Ocean contained the ocean’s
lowest d13C values, suggesting a local accumulation of remineralized C and/or poor ventilation
(24). Spero and Lea (25) argued that during the
early part of the last deglaciation, Southern
Ocean sea-ice retreat (23), combined with increased deep convection and northward Ekman
transport, imparted transient low d13C values to
AAIW and Subantarctic Mode Waters and that
this signal was recorded by deep-dwelling
planktonic foraminifera in various ocean basins,
including the eastern tropical North Pacific.
These waters should also have carried a low
D14C signature, and we suggest that this is the
signal we observe off the coast of Baja California. Today, AAIW is barely traceable as a
distinct water mass north of the equator in the
eastern tropical Pacific (26). We argue that the
northward penetration of AAIW during deglaciation was greater than it is today (27) and
was at times fed by extremely 14C-depleted
waters sourced from the abyss by deep overturning in the Southern Ocean. Sea-ice retreat
could have allowed the upwelled deep waters to
gain buoyancy from precipitation, converting
some fraction of these waters into AAIW without substantial mixing with warmer thermocline
waters (28), which would otherwise dampen the
D14C signal. There is evidence that vertical
stratification of the North Pacific also varied on
an Antarctic climate schedule (29), so northern
deep waters may have supplemented the supply
of aged C to the Baja California site. However,
1458
Fig. 2. Baja California intermediate-water D14C during the last deglaciation (red), compared with
atmospheric D14C (dark green, light blue, and dark blue) (1, 3, 17), and atmospheric CO2 from Antarctica
Dome C (6) placed on the GISP2 time scale (black), as discussed in the text. Vertical dashed gray lines
show the ages of Bølling-Allerød (B-A) and Younger Dryas (YD) boundaries, based on the GISP2 d18O
record, and the start of Heinrich event 1 (H1), based on the 231Pa/230Th record from Bermuda Rise (34).
The ACR is contemporaneous with the Bølling-Allerød. ppmv, parts per million by volume.
the interaction of strong circumpolar winds with
bathymetry in the Southern Ocean provides for
much more effective vertical pumping (30) than
do conditions in the North Pacific, and therefore
southern sources probably dominated the D14C
changes in our record.
North-south teleconnections. Numerous observational and modeling studies indicate an
inverse relationship between Antarctic and
North Atlantic temperature variations that may
be due to altered interhemispheric ocean-heat
transport and/or opposing local deep-water
formation histories (28, 31–33). Insofar as
D14C of the intermediate-depth Pacific provides
an inverse proxy for the strength of deep convection in the Southern Ocean, our results provide strong evidence for tight, inverse coupling
of deep-water formation between hemispheres.
This is clearly demonstrated by the covariation
of intermediate-depth Pacific D14C and the
formation history of North Atlantic Deep Water
(NADW) [or its glacial analog, Glacial North
Atlantic Intermediate Water (GNAIW)], based
on measurements of 231Pa/230Th from sediments
in the western North Atlantic (34) (Fig. 3C). The
near-cessation of 231Pa export beginning just
after 18,000 yr B.P. records a collapse of
GNAIW that has been linked to a massive discharge of glacial ice and fresh water to the North
Atlantic, known as Heinrich event 1. After a
recovery during the Bølling-Allerød warm
phase, another marked weakening of NADW/
GNAIW is documented during the Younger
8 JUNE 2007
VOL 316
SCIENCE
Dryas cold period, probably also triggered by a
discharge of glacial meltwater (34). Both periods
of NADW/GNAIW reduction were times of
intermediate-depth Pacific D14C decline and
atmospheric CO2 rise.
It is often difficult to identify triggers in a
tightly coupled system, but the relationships
described above suggest that the Antarctic climate schedule may have been paced by ice sheet
and meltwater forcing around the North Atlantic.
Support for this conclusion comes from our
observation that although major inflections in the
D14C and 231Pa/230Th records in Fig. 3 were
almost exactly synchronous, the large D14C
decrease (i.e., Southern Ocean ventilation increase) during Heinrich event 1 occurred more
gradually than did the associated decrease in
GNAIW export. This relationship is consistent
with relatively slow circum-Antarctic warming
due to anomalous ocean-heat transport (33),
leading to sea-ice retreat (23) [and possibly
poleward-shifting westerlies (8)] and, consequently, a progressive increase in deep overturning of the Southern Ocean. Alternatively,
Southern Ocean overturning may have been
instigated by NADW/GNAIW reductions
through the requirement (35) that global rates
of deep-water formation balance global deep
upwelling, which is forced mainly by winds and
tides. The abrupt rise in atmospheric D14C at the
start of the Younger Dryas [+80‰ in just 180
years (36)] (Fig. 2) may record the time elapsed
before the Southern Ocean could begin respond-
www.sciencemag.org
Downloaded from www.sciencemag.org on June 7, 2007
RESEARCH ARTICLES
Fig. 3. Southern and northern ocean-atmosphere changes
during the last deglaciation,
compared with intermediatewater D14C. (A) Atmospheric
CO2 (black) and ice core
deuterium (dD) temperature
proxy (light blue) from Antarctica Dome C (6), placed on
the GISP2 time scale. (B) Baja
California intermediate-water
D14C. (C) Inverted decaycorrected excess 231Pa/230Th
in Bermuda Rise sediments,
based on two methods to
calculate excess (green and
purple) (34). The horizontal
dashed line shows the watercolumn production ratio
for these isotopes (0.093);
lower values are primarily
due to Pa export by vigorous NADW. Vertical dashed
lines show the ages of climatic boundaries, as in Fig.
2. Error bars are based on
compound uncertainties in
radiocarbon ages and calendar ages.
ing to reduced NADW formation, leading to the
brief absence of deep-ocean sinks for 14C (32).
Recent work shows that deep North Atlantic
radiocarbon activities also increased abruptly
during the Bølling-Allerød because of renewed
formation of NADW and that they temporarily
decreased during the Younger Dryas, when
NADW formation was again briefly restricted
(37, 38). In light of our new record, this pattern
of deep-ocean change was probably limited to
the North Atlantic (arising from deep circulation
changes within the basin), whereas the
intermediate-depth North Pacific record tracks
the overall redistribution of C from the deep
ocean to the atmosphere.
Implications for the deep C reservoir. Finally,
our results bear on recent questions concerning
reconstructed rates of atmospheric D14C decline
during deglaciation and implied 14C aging of
glacial deep waters. Although subject to uncertainties (19), decay projection of our first
deglacial D14C minimum back to the surface
ocean (39) gives an apparent ventilation age of
~4000 years, implying that the inferred deep
Southern Ocean source waters were at least that
old. This is broadly consistent with the
minimum deep-ocean age estimated from the
atmospheric record, assuming that the old
reservoir filled half of the ocean’s volume (40).
Ages of up to 5000 years have been reported for
glacial deep waters near New Zealand (41), but
more northerly sites in the Pacific show little difference from today, at least at depths shallower
than ~2 km (40). We infer that the greatest 14C depletion of the glacial deep ocean was probably
concentrated in the Southern Ocean region (and
deepest Pacific), coincident with the highest densities (22) and lowest d13C values (24).
References and Notes
1. R. G. Fairbanks et al., Quat. Sci. Rev. 24, 1781 (2005).
2. K. Hughen et al., Science 303, 202 (2004).
3. K. Hughen, J. Southon, S. J. Lehman, C. Bertrand,
J. Turnbull, Quat. Sci. Rev. 25, 3216 (2006).
4. C. Laj et al., Earth Planet. Sci. Lett. 200, 177 (2002).
5. R. Muscheler et al., Earth Planet. Sci. Lett. 219, 325 (2004).
6. E. Monnin et al., Science 291, 112 (2001).
7. D. M. Sigman, E. A. Boyle, Nature 407, 859 (2000).
8. J. R. Toggweiler, J. L. Russell, S. R. Carson,
Paleoceanography 21, PA2005 (2006).
9. B. B. Stephens, R. F. Keeling, Nature 404, 171 (2000).
10. A. van Geen et al., Paleoceanography 18, 1098 (2003).
11. J. D. Ortiz et al., Geology 32, 521 (2004).
12. R. J. Behl, J. P. Kennett, Nature 379, 243 (1996).
13. Y. J. Wang et al., Science 294, 2345 (2001).
14. W. E. Dean, Y. Zheng, J. D. Ortiz, A. van Geen,
Paleoceanography 21, PA4209 (2006).
www.sciencemag.org
SCIENCE
VOL 316
15. P. M. Grootes, M. Stuiver, J. Geophys. Res. 102, 26455
(1997).
16. Materials and methods are available as supporting material
on Science Online. In the supporting material, we also
present our D14C values calculated with the original GISP2
chronology (15) and compare them to the original Cariaco
record (2), based on correlation to GISP2 (fig. S3).
17. P. J. Reimer et al., Radiocarbon 46, 1029 (2004).
18. G. Ostlund, H. Craig, W. S. Broecker, D. Spencer,
GEOSECS Atlantic, Pacific and Indian Ocean Expeditions,
vol. 7, Shorebased Data and Graphics (U.S. Government
Printing Office, Washington, DC, 1987).
19. Throughout most of this paper, we discuss our D14C
values with respect to the contemporaneous atmosphere,
rather than calculating ventilation ages by decay
projection (39) because of large uncertainties in initial
surface-ocean reservoir ages and the possibility of
multiple intermediate- and deep-water sources with
different ventilation histories.
20. U. Mikolajewicz, T. J. Crowley, A. Schiller, R. Voss, Nature
387, 384 (1997).
21. The deglacial injection of low-14C intermediate waters
along the Baja California margin did not noticeably
overprint the DSR or other O2-sensitive productivity
proxies recorded in GC31/PC08. Although D14C and O2
are broadly correlated in the modern ocean, they can
also be decoupled (18) through air-sea exchange. Hence,
deglacial intermediate waters were very depleted in
D14C but apparently not drastically depleted in O2.
22. J. F. Adkins, K. McIntyre, D. P. Schrag, Science 298, 1769
(2002).
23. R. Gersonde, X. Crosta, A. Abelmann, L. Armand, Quat.
Sci. Rev. 24, 869 (2005).
24. U. S. Ninnemann, C. D. Charles, Earth Planet. Sci. Lett.
201, 383 (2002).
25. H. J. Spero, D. W. Lea, Science 296, 522 (2002).
26. P. C. Fiedler, L. D. Talley, Prog. Oceanogr. 69, 143 (2006).
27. S. Schulte, F. Rostek, E. Bard, J. Rullkötter, O. Marchal,
Earth Planet. Sci. Lett. 173, 205 (1999).
28. R. F. Keeling, B. B. Stephens, Paleoceanography 16, 112
(2001).
29. S. L. Jaccard et al., Science 308, 1003 (2005).
30. J. R. Toggweiler, B. Samuels, Deep-Sea Res. 42, 477 (1995).
31. T. J. Crowley, Paleoceanography 7, 489 (1992).
32. W. S. Broecker, Paleoceanography 13, 119 (1998).
33. R. Knutti, J. Flückiger, T. F. Stocker, A. Timmermann,
Nature 430, 851 (2004).
34. J. F. McManus, R. Francois, J.-M. Gherardi, L. D. Keigwin,
S. Brown-Leger, Nature 428, 834 (2004).
35. W. Munk, C. Wunsch, Deep-Sea Res. 45, 1977 (1998).
36. K. A. Hughen, J. R. Southon, S. J. Lehman, J. T. Overpeck,
Science 290, 1951 (2000).
37. L. C. Skinner, N. J. Shackleton, Paleoceanography 19,
PA2005 (2004).
38. L. F. Robinson et al., Science 310, 1469 (2005).
39. J. F. Adkins, E. A. Boyle, Paleoceanography 12, 337 (1997).
40. W. Broecker et al., Science 306, 1169 (2004).
41. E. L. Sikes, C. R. Samson, T. P. Guilderson, W. R. Howard,
Nature 405, 555 (2000).
42. We thank D. Lopez, J. Turnbull, and C. Wolak for laboratory
assistance; J. Southon for accelerator mass spectrometry
analyses; A. Pearson for providing an unpublished seawater
D14C measurement; and T. Blunier for providing Dome
C EDC 3 time scale and for valuable discussions on
synchronization of the different ice cores. This manuscript
was improved by comments from R. Keeling and two
anonymous reviewers. Support was provided by NSF grants
OCE-9809026 and OCE-0214221.
Downloaded from www.sciencemag.org on June 7, 2007
RESEARCH ARTICLES
Supporting Online Material
www.sciencemag.org/cgi/content/full/1138679/DC1
Materials and Methods
SOM Text
Figs. S1 to S3
Tables S1 and S2
References
11 December 2006; accepted 27 April 2007
Published online 10 May 2007;
10.1126/science.1138679
Include this information when citing this paper.
8 JUNE 2007
1459
www.sciencemag.org/cgi/content/full/1138679/DC1
Supporting Online Material for
Marine Radiocarbon Evidence for the Mechanism of Deglacial
Atmospheric CO2 Rise
T. M. Marchitto,* S. J. Lehman, J. D. Ortiz, J. Flückiger, A. van Geen
*To whom correspondence should be addressed. E-mail: [email protected]
Published 10 May 2007 on Science Express
DOI: 10.1126/science.1138679
This PDF file includes:
Materials and Methods
SOM Text
Figs. S1 to S3
Tables S1 and S2
References
Supporting Online Material
Methods
Bulk sediment samples were washed using a sodium hexametaphosphate solution and
picked for benthic foraminifera using the >250 µm size fraction, supplemented with 150250 µm specimens where necessary. Toward the end of this study, picked foraminifera
were also briefly sonicated in methanol (Table S1). Foraminiferal samples were
hydrolyzed with H3PO4 and reduced to graphite using an Fe catalyst in the presence of
H2. The 19 published radiocarbon samples (S1) were graphitized at the National Ocean
Sciences AMS Facility in Woods Hole (NOSAMS), while the 31 new samples were
graphitized at the INSTAAR Laboratory for AMS Radiocarbon Preparation and Research
at the University of Colorado, Boulder (NSRL). All published radiocarbon data (S1) were
analyzed at NOSAMS, while all new data were analyzed at the Keck Carbon Cycle AMS
Facility at the University of California, Irvine (KCCAMS). δ13C corrections were based
on off-line measurement for NOSAMS results, and on measurement within the AMS for
KCCAMS results. KCCAMS δ13C values are not suitable for paleoceanographic
interpretation, and are therefore not reported. Age-corrected ∆14C calculations were based
on the conventions of Stuiver and Polach (S2). ∆14C error bars were calculated by
compounding radiocarbon age errors with estimated calendar age errors listed in Table
S1. Calendar age errors depend solely on the estimated precision of our picks with
respect to the GISP2-Hulu chronology, and not on any errors in the GISP2-Hulu
chronology itself. The latter would affect both our record and the Cariaco Basin
atmospheric ∆14C record similarly.
Hulu Cave adjustments to GISP2 age model
In order to compare PC08 results with the most recent and most consistent
atmospheric ∆14C reconstructions (S3-S5), the correlation between PC08 reflectance and
GISP2 δ18O (S6) (used to derive calendar ages for calculating PC08 ∆14C) requires that
the GISP2 chronology be consistent with the Hulu Cave chronology (S7), as outlined in
the main text. Younger than ~20-25 kyr BP the Hulu Cave and GISP2 chronologies
appear to be consistent with each other, so no corrections need to be applied to the more
recent parts of the GISP2 age model (Fig. S1). GISP2 also agrees with the latest NGRIP
chronology (GICC05) to within 250 yr over this interval (S8). We adjusted the GISP2
chronology older than 23.4 kyr BP (the start of Interstadial 2) by linearly interpolating
between three Hulu Cave tie-points corresponding to major interstadial warmings in
Greenland. The resulting GISP2 age model appears to agree with Hulu Cave to within a
few hundred years throughout the past 50 kyr (Fig. S1). The resulting age-depth
relationship for PC08 is shown in Fig. S2A.
Methane synchronization of Dome C and GISP2
As outlined above and in the main text, the GISP2, Hulu Cave, and coral
chronologies are similar over the last ~20-25 kyr. However, these chronologies differ
from the glaciological age model on which the EPICA Dome C CO2 and δD records of
Monnin et al. (S9) were originally presented. To obtain a timescale that is consistent with
the various ∆14C reconstructions, we have used CH4 records from the Dome C, GISP2,
and GRIP ice cores (S9-S11) to derive a GISP2 synchronized gas age scale for Dome C.
Seven tie-points (Table S2) were defined based on the comparison of the Dome C CH4
record (on the EDC3 time scale (S12, S13)) and the GISP2 and GRIP CH4 records (on the
GISP2 time scale (S10)). Between the tie-points the EDC3 gas age scale was linearly
interpolated to GISP2 ages. The associated ice age scale for Dome C was derived from
the synchronized gas age scale described above and the delta ages (gas age to ice age
differences in the ice core) taken from the EDC3 time scale (S12, S13).
Sedimentary artifacts
The two deglacial intervals of depleted radiocarbon activity that we report appear, in
part, as large age plateaus in the conventional radiocarbon age:depth relationship (Table
S1). During the first (earlier) event, radiocarbon age is nearly constant over 90 cm of core
depth, and during the second event, radiocarbon age is nearly constant over at least 45
cm. Near-instantaneous massive inputs of sediment, namely slumps, could theoretically
produce age plateaus. However, such events would necessarily produce large kinked
offsets in the calendar age:depth relationship, which are clearly not observed (Fig. S2).
Additionally, slumps are much more likely to produce age reversals than plateaus.
Radiocarbon age plateaus could also theoretically be created during intervals of very
intense bioturbation, such that age is homogenized over some depth interval. However, it
is unreasonable to postulate a complete, simultaneous bioturbational homogenization of
the upper 90 cm (or 45 cm) of the sediment column, especially in this low-O2
environment. The entire MC19/GC31/PC08 record is dominated by low-O2 benthic
foraminiferal taxa (mainly Bolivina and Uvigerina), and sedimentary Mo remains
enriched by ~4-8 times crustal values during deglaciation, indicative of shallow pore
water sulfide and suboxic bottom waters (S14). Finally, very intense bioturbation would
homogenize other paleoclimate proxies in this core. Notably, our second age plateau
crosses the Allerød-Younger Dryas boundary in GC31/PC08, which is marked by a sharp
reflectance change that is inferred to record a shift in surface ocean productivity (S15).
We therefore conclude decisively that the 14C age-depth patterns in our record reflect
changes in intermediate water 14C activity and not sedimentary processes.
Alternate PC08 age model
We have also calculated Baja California ∆14C based on correlation to GISP2 using the
original layer-counted age model of Meese et al. (S16) (Fig. S3). This record only differs
from our preferred GISP2-Hulu age model prior to 23.4 kyr BP. In Fig. S3 we compare
this alternate record to the original Cariaco Basin ∆14C record based on correlation to
GISP2 (S17). It is clear that our conclusions are not affected by the choice of age model,
though the agreement between Cariaco Basin and coral ∆14C is better in Fig. 1 than in
Fig. S3 (S5). The PC08 age-depth relationship for the alternate age model is shown in
Fig. S2B.
2
Fig. S1. Derivation of GISP2-Hulu Cave composite age model. (A) Diffuse spectral
reflectance factor 3 from Baja California composite sediment core MV99-GC31/PC08,
plotted versus depth (S15). Gray lines show tie-points to Greenland record used to derive
calendar age model. (B) δ18O of Greenland ice core GISP2 (S6) on provisional revised
timescale (black). New timescale deviates from original timescale (green) older than 23.4
kyr BP (line labeled G). New timescale is based on linear interpolation between point G
and three tie-points whose ages are derived from U-Th-dated Hulu Cave (lines labeled
H). (C) Hulu Cave δ18O from three different stalagmites (S7). Gray lines indicate tiepoints used to adjust GISP2 chronology.
3
Fig. S2. (A) Age-depth relationship for MV99-MC19/GC31/PC08 based on correlation to
GISP2 δ18O with Hulu Cave age adjustment. Red points are tie-points and gray points
represent radiocarbon measurements. The unusually constant sedimentation rate allows
for relatively precise calendar age assignments between tie-points. (B) Same as (A)
except using alternate age model as in Fig. S3.
4
Fig. S3. Baja California ∆14C record based on alternate age model. (A) Diffuse spectral
reflectance factor 3 from Baja California composite sediment core MV99-GC31/PC08,
plotted versus depth (top axis) (S15). Gray lines show tie-points to Greenland record used
to derive alternate calendar age model. (B) δ18O of Greenland ice core GISP2 on original
timescale (bottom axis) (S6, S16). (C) Atmospheric radiocarbon activities based on tree
rings, planktonic foraminifera from Cariaco Basin varve-counted sediments, and U-Thdated corals (dark green) (S3); additional recent coral measurements (cyan) (S4); and
planktonic foraminifera from Cariaco Basin using age model derived from reflectance
correlation to GISP2 (blue) (S17). Red points show intermediate water activities from
benthic foraminifera in MC19/GC31/PC08. Yellow point is estimate for modern bottom
waters at this site.
5
Sample
MC19-10 cm
MC19-25 cm
GC31-3-2 cm
MC19-40 cm
GC31-3-22.5 cm
GC31-3-50.5 cm
GC31-3-70 cm
GC31-3-100.5 cm
GC31-3-123 cm
GC31-2-0.5 cm
GC31-2-50.5 cm
GC31-2-100.5 cm
GC31-1-8.5 cm
PC08-9-45 cm
PC08-9-60 cm
GC31-1-32.5 cm
PC08-9-85 cm
GC31-1-58.5 cm
PC08-9-110 cm
PC08-9-130 cm
GC31-1-94.5 cm
PC08-9-141 cm
PC08-9-150 cm
PC08-8-6.5 cm
PC08-8-16.5 cm
PC08-8-25.5 cm
PC08-8-35 cm
Taxa
mixed benthics
mixed benthics
mixed benthics
mixed benthics
mixed benthics
Bolivina spp.
mixed benthics
Bolivina spp.
mixed benthics
Bolivina spp.
Bolivina spp.
Bolivina spp.
Bolivina spp.
mixed benthics
mixed benthics
mixed benthics
mixed benthics
Bolivina spp.
mixed benthics
mixed benthics
Bolivina spp.
Uvigerina spp.
mixed benthics
Uvigerina spp.
Uvigerina spp.
Uvigerina spp.
mixed benthics
Composite
depth (m)
0.100
0.250
0.270
0.400
0.475
0.755
0.950
1.255
1.480
1.755
2.255
2.755
3.255
3.260
3.410
3.495
3.660
3.755
3.910
4.110
4.115
4.220
4.310
4.375
4.475
4.565
4.660
C age
(kyr BP)
1.720
2.050
2.050
2.320
2.230
3.030
3.690
3.840
5.130
5.810
7.190
8.980
10.050
10.460
10.845
11.600
13.380
13.500
13.530
13.420
13.650
14.285
13.370
14.485
15.755
15.850
16.505
14
Table S1. Core MC19/GC31/PC08 radiocarbon data.
C age
error
(kyr)
0.030
0.035
0.030
0.035
0.035
0.040
0.045
0.050
0.060
0.040
0.050
0.060
0.410
0.030
0.030
0.070
0.035
0.070
0.030
0.025
0.150
0.035
0.030
0.035
0.040
0.040
0.040
14
Calendar
age
(kyr BP)
0.32
0.81
0.87
1.29
1.54
2.44
3.08
4.07
4.79
5.69
7.31
9.01
10.79
10.81
11.32
11.61
12.12
12.41
12.88
13.57
13.58
13.94
14.20
14.37
14.65
14.96
15.30
Calendar
age error
(kyr)
0.1
0.2
0.3
0.4
0.4
0.4
0.4
0.4
0.4
0.4
0.3
0.3
0.2
0.2
0.2
0.1
0.2
0.2
0.1
0.1
0.3
0.1
0.2
0.2
0.1
0.2
0.3
∆14C
(per
mil)
-161
-146
-139
-124
-88
-78
-84
14
-57
-35
-11
-27
56
6
20
-39
-181
-164
-118
-29
-54
-88
55
-62
-172
-150
-184
-1.43
-1.49
-1.49
δ13C
(per
mil)
-0.61
-0.69
-0.86
-0.56
-0.91
nm
-0.60
nm
-0.53
nm
nm
-0.90
-1.15
Accession #
OS-33198
OS-33199
OS-33201
OS-33200
OS-25612
OS-22946
OS-33202
OS-22947
OS-33203
OS-22948
OS-22949
OS-22955
OS-23513
CURL-8750
CURL-8752
OS-25611
CURL-8751
OS-22956
CURL-8444
CURL-8445
OS-22957
CURL-8746
CURL-8446
CURL-8721
CURL-8726
CURL-8720
CURL-8447
Son?
no
no
no
no
no
no
no
no
no
no
no
no
no
yes
yes
no
yes
no
no
no
no
yes
no
yes
yes
yes
no
6
Reference
1
1
1
1
1
1
1
1
1
1
1
1
1
this study
this study
1
this study
1
this study
this study
1
this study
this study
this study
this study
this study
this study
7
PC08-8-45.5 cm
Uvigerina spp.
4.765
16.785
0.045
15.67
0.3
-176
CURL-8724 yes
this study
PC08-8-55.5 cm
Uvigerina spp.
4.865
16.665
0.040
16.02
0.3
-127
CURL-8729 yes
this study
PC08-8-65.5 cm
Uvigerina spp.
4.965
16.425
0.045
16.37
0.3
-62
CURL-8744 yes
this study
PC08-8-75.5 cm
Uvigerina spp.
5.065
16.390
0.040
16.73
0.3
-17
CURL-8742 yes
this study
PC08-8-85.5 cm
Uvigerina spp.
5.165
16.425
0.050
17.08
0.3
22
CURL-8743 yes
this study
PC08-8-95.5 cm
Uvigerina spp.
5.265
16.185
0.045
17.43
0.3
98
CURL-8749 yes
this study
PC08-8-105.5 cm
Uvigerina spp.
5.365
16.235
0.040
17.78
0.3
139
CURL-8748 yes
this study
PC08-8-116 cm
Uvigerina spp.
5.470
16.810
0.040
18.12
0.3
105
CURL-8745 yes
this study
PC08-8-125 cm
mixed benthics
5.560
16.680
0.040
18.41
0.3
162
CURL-8448 no
this study
PC08-7-60.5 cm
Uvigerina spp.
6.415
19.720
0.060
21.14
0.2
108
CURL-8728 yes
this study
PC08-7-80.5 cm
Uvigerina spp.
6.615
19.640
0.060
21.77
0.3
208
CURL-8722 yes
this study
PC08-7-99.75 cm
mixed benthics
6.808
19.650
0.080
22.39
0.3
299
-0.81
OS-33204
no
1
PC08-7-132.5 cm
Uvigerina spp.
7.135
21.170
0.070
23.43
0.1
220
CURL-8727 yes
this study
PC08-6-20.5 cm
Uvigerina spp.
7.585
22.260
0.070
25.15
0.3
311
CURL-8725 yes
this study
PC08-6-118 cm
mixed benthics
8.560
25.500
0.170
29.00
0.2
396
-0.57
OS-33205
no
1
PC08-5-25 cm
mixed benthics
9.010
27.200
0.130
30.73
0.2
394
CURL-8449 no
this study
PC08-5-95 cm
mixed benthics
9.710
29.230
0.160
33.11
0.2
443
CURL-8450 no
this study
PC08-5-130 cm
Uvigerina spp.
10.060
30.830
0.170
34.55
0.2
407
CURL-7188 no
this study
PC08-5-145 cm
Uvigerina spp.
10.210
30.460
0.190
35.11
0.2
577
CURL-7189 no
this study
PC08-5-145 cm
Bolivina spp.
10.210
31.370
0.180
35.11
0.2
408
CURL-7192 no
this study
PC08-4-15 cm
mixed benthics
10.410
31.200
0.280
35.89
0.2
582
-0.72
OS-33206
no
1
PC08-4-15 cm
Uvigerina spp.
10.410
31.750
0.190
35.89
0.2
477
CURL-7187 no
this study
PC08-4-65 cm
Bolivina spp.
10.910
34.400
0.260
37.92
0.2
356
CURL-7193 no
this study
Composite depth is relative to the top of MC19, as in (S1). Calendar age errors are based on estimated precision of tie-points and distance of samples from tiepoints. δ13C data are only reported for NOSAMS measurements; ‘nm’ indicates not measured. Column labeled ‘Son?’ indicates whether or not radiocarbon
samples were sonicated in methanol prior to graphitization.
Table S2. Tie-points between the EPICA Dome C EDC3 and GISP2 time scales.
EDC3
gas age
(yr BP)
0
8200
11400
12400
14110
16700
21860
GISP2
gas age
(yr BP)
0
8200
11620
12750
14810
17600
22950
estimated
uncertainty
(yr)
0
25
25
100
150
200
300
event description
8.2 ka BP event
YD termination
Onset of YD
Onset of B-A
First deglacial CH4 rise
CH4 peak in GRIP and Dome C
8
SOM References
S1.
S2.
S3.
S4.
S5.
S6.
S7.
S8.
S9.
S10.
S11.
S12.
S13.
S14.
S15.
S16.
S17.
A. van Geen et al., Paleoceanography 18, doi:10.1029/2003PA000911 (2003).
M. Stuiver, H. A. Polach, Radiocarbon 19, 355 (1977).
P. J. Reimer, et al., Radiocarbon 46, 1029 (2004).
R. G. Fairbanks et al., Quaternary Science Reviews 24, 1781 (2005).
K. Hughen, J. Southon, S. J. Lehman, C. Bertrand, J. Turnbull, Quaternary
Science Reviews, 25, 3216 (2006).
P. M. Grootes, M. Stuiver, Journal of Geophysical Research 102, 26455 (1997).
Y. J. Wang et al., Science 294, 2345 (2001).
A. Svensson et al., Quaternary Science Reviews, 25, 3258 (2006).
E. Monnin et al., Science 291, 112 (2001).
T. Blunier, E. Brook, Science 291, 109 (2001).
J. Flückiger et al., Global Biogeochemical Cycles 16, 1010 (2002).
F. Parrenin et al., Climate of the Past Discussions 3, 19 (2007).
L. Loulergue et al., Climate of the Past Discussions 3, 425 (2007).
W. E. Dean, Y. Zheng, J. D. Ortiz, A. van Geen, Paleoceanography 21,
doi:10.1029/2005PA001239 (2006).
J. D. Ortiz et al., Geology 32, 521 (2004).
D. A. Meese et al., Journal of Geophysical Research 102, 26411 (1997).
K. Hughen et al., Science 303, 202 (2004).
9
PERSPECTIVES
further work on the stress accumulation and
dissipation in the lithosphere during subduction is necessary to understand the faulting
mechanism causing seismicity in double WBZ
or even triple WBZ, as proposed for some
regions beneath Japan (6).
Brudzinski et al. show that, as the number
of seismological stations and the availability
of digital seismic traces increases, the global
earthquake catalog will become accurate
enough to delineate the fine structure of seismicity (on the order of a few kilometers) on a
global scale. This accuracy will increase in the
near future as a result of large seismological
observation initiatives like EarthScope in
the United States or the NERIES program
(Network of Research Infrastructures for
European Seismology) in Europe, which will
make more high-quality digital data readily
available for seismologists worldwide.
References
1. M. R. Brudzinski, C. H. Thurber, B. R. Hacker, E. R.
Engdahl, Science 316, 1472 (2007).
2. A. Hasegawa, N. Umino, A. Takagi, Tectonophysics 47, 43
(1978).
3. A. Rietbrock, F. Waldhauser, Geophys. Res. Lett. 31,
10.1029/2004GL019610 (2004).
4. S. Kirby, Rev. Geophys. 33, 287 (1995).
5. B. R. Hacker, G. A. Abers, S. M. Peacock, J. Geophys. Res.
108, 2029 (2003).
6. T. Igarashi, T. Matsuzawa, N. Umino, A. Hasegawa,
J. Geophys. Res. 106, 2177 (2001).
10.1126/science.1141921
ATMOSPHERE
Results from a sediment core provide insights
into ocean circulation changes during the last
deglaciation.
Deglaciation Mysteries
Ralph F. Keeling
etween 19,000 and 11,000
waters, which is a measure of
Glacial state
SOUTH
NORTH
Cap
0m
years ago, the Earth emthe deep-water ventilation rate.
Core site
erged from the last glacial
This technique, applied to
period. During this deglaciation,
numerous sediment cores, has
–2800 m
Poorly ventilated deep water
the carbon dioxide (CO2) concenthus far mainly yielded the unre14
Low C
tration in the atmosphere rose from
markable result that the ventila–4000 m
180 to 265 parts per million (ppm).
tion rate of the glacial ocean was
Over the same period, the radiovery similar to that of today’s
CO2 release
Early deglacial state
0m
carbon content of the CO2 fell by
ocean, at least down to a depth
Low 14C intermediate water
~35%. A simple but unproven exof ~2800 m (2). Thus, if there
planation for both changes is an
was a major change in ocean
–2800 m
increase in the rate at which the
ventilation during deglaciation,
Poorly ventilated deep water
ocean’s subsurface waters were
this change must have occurred
–4000 m
renewed by exchange with aerated
in ocean waters below that depth.
surface waters—a process known
However, despite some tanas ventilation. A ventilation in- Deep-water ventilation. This cross section of the Pacific Ocean shows how poorly talizing results (3), no general
crease could increase atmospheric ventilated water may have been delivered to intermediate depths during deglacia- picture has emerged for waters
tion, as suggested by Marchitto et al. (Top) Ventilation of the deep ocean by sinkCO2 concentration by releasing
below 2800 m, because of mething around Antarctica was partially suppressed by a cap formed by sea ice or a layer
excess CO2 that had accumulated of low-salinity water. (Bottom) This cap was removed during early deglacial warm- odological difficulties related
in subsurface waters by the decom- ing, exposing upwelled deep waters to the atmosphere, releasing radiocarbon- to the low sedimentation rates
position of sinking detritus. On depleted CO . The density of the poorly ventilated waters was reduced by freshen- that typically characterize cores
2
page 1456 in this issue, Marchitto ing and warming. With reduced density, the water could spread widely at interme- from these depths.
et al. (1) bolster the case for such diate depths, displacing waters of similar density.
Marchitto et al. focus on a
a ventilation increase and offer insediment core recently hauled
of dissolved carbon. Because of radioactive up off Baja California from a depth of 700 m,
sight into how the increase may have occurred.
To track changes in past ventilation, most decay, waters that are more isolated from the seemingly too shallow for studying deep venresearchers have turned to measurements of atmosphere (that is, more poorly ventilated) tilation. The core contains bands correspondthe radiocarbon (14C) content of shells of have lower 14C/12C ratios, as do the shells that ing to a set of millennial climate oscillations
foraminifera, a ubiquitous marine microor- grow in these waters. The radiocarbon age of a first discovered in ice cores from Greenland,
ganism. Radiocarbon is produced naturally in fossil shell therefore reflects the age of the allowing absolute dates to be fixed within the
the upper atmosphere by cosmic rays and shell plus the age of the water in which it lived. core. Using these dates, the authors correct the
spreads through the ocean as part of the pool By subtracting the radiocarbon ages of sur- 14C/12C ratios of benthic foraminifera for
face-dwelling (planktonic) and bottom- radioactive decay, thus establishing the origidwelling (benthic) foraminifera, picked from nal 14C/12C ratio of the water in which the
The author is at the Scripps Institution of Oceanography,
the same layer of a sediment core, it is possible foraminifera lived. The technique does not
University of California, San Diego, CA 92093, USA.
to estimate age difference of surface and deep require 14C/12C measurements on planktonic
E-mail: [email protected]
B
1440
8 JUNE 2007
VOL 316
SCIENCE
Published by AAAS
www.sciencemag.org
Downloaded from www.sciencemag.org on June 7, 2007
Brudzinski et al. have now found a direct
correlation whereby older oceanic plates
show a greater distance between regions of
seismicity. They conclude, based on thermalpetrological models developed by Hacker et
al. (5), that dehydration of the mineral antigorite is responsible for the seismicity in the
lower layer of double WBZs.
Although the detailed geophysical explanation presented by Brudzinski et al. for the
double seismic zone might be debatable, they
observe that double WBZs are the rule and not
the exception during subduction of oceanic
lithosphere. This provides an important new
constraint for all models developed to explain
the occurrence of WBZ seismicity. However,
foraminifera, which are subject to potentially
large systematic errors.
The results show two periods during
deglaciation when the bottom water at their site
had unexpectedly low 14C/12C ratios. The water
was so old that it must have been delivered to
the site by upwelling from greater depth, presumably from below 2800 m. The oldest waters
found by Marchitto et al. have an age of ~4000
years. For comparison, the oldest waters in the
modern ocean have an age of ~2300 years.
The study provides the strongest evidence
to date that the glacial ocean contained some
very poorly ventilated water somewhere in its
depths. The low-14C periods coincide with
the periods when atmospheric radiocarbon
decreased and atmospheric CO2 increased
most rapidly during deglaciation. The results
are thus a convincing fingerprint of a process
that flushed excess carbon from an isolated
deep reservoir toward the surface, thereby
driving the atmospheric changes.
Today, waters below 2800 m are ventilated
by two routes. One involves the sinking of aerated surface waters in the North Atlantic, the
other sinking of such waters near Antarctica.
During the last glacial period, both routes
probably weakened, with the southern route
possibly influenced by sea ice or surface
freshening (see the figure, top panel). During
glacial times, the deep ocean would thus have
been less ventilated than it is today.
But how could low-14C waters get to
Marchitto et al.’s core site during deglaciation?
Much of the upwelling of deep water occurs
today around Antarctica, resulting in the formation of Antarctic Intermediate Water, a lowsalinity water mass that spreads northward at
intermediate depths. Marchitto et al. hypothesize that a similar process occurred during
deglaciation, allowing upwelled water to
spread northward to their site (see the figure,
bottom panel). However, the evidence for this
southern pathway is circumstantial.
The results help to reconcile the reconstructed trends in atmospheric radiocarbon
with the estimated trends in the production of
radiocarbon by cosmic rays—a comparison
that seems to demand an increase in ocean
ventilation during deglaciation (4). They support theories that attribute the bulk of the glacial-interglacial CO2 change to changes in
ocean ventilation (5, 6).
The study also provides support for a theory for how the glacial ocean differed from
today’s ocean as a result of the cooling of
deep waters to nearly the freezing point.
Cooling to this extent is expected to allow the
salty brine that is released during sea ice formation to accumulate more easily in the deep
ocean (7). This idea is supported by sediment
pore-water studies (8). By blocking the input
of fresh water from precipitation, sea ice
could also reduce the conversion of upwelled
deep water into low-salinity Antarctic Intermediate Water (7). A strengthening of intermediate-water formation during deglaciation
is consistent with a breakdown of this state
caused by warming.
The study nevertheless leaves the skeptics
MOLECULAR BIOLOGY
with arrows in their quiver. Marchitto et al.’s
low-14C waters are so old that they start to
stretch credibility, especially considering that
the deep reservoir from which the water was
drawn must have been even older. (This follows because some mixing with younger
water would unavoidably have occurred during upwelling and transit to the site.) How
could prior studies have overlooked deep
waters this old?
If Marchitto et al.’s interpretation is correct, evidence for old water at intermediate
depths should be present throughout the South
Pacific in sediments of the appropriate age
and depth. If subsequent work supports the
findings, we may look back at this study as a
key turning point in the quest to understand
glacial and interglacial CO2 changes.
References
1. T. M. Marchitto, S. J. Lehman, J. D. Ortiz, J. Flückiger,
A. van Geen, Science 316, 1456 (2007); published
online 10 May 2007 (10.1126/science. 1138679).
2. W. Broecker et al., Paleoceanography 22, PA2206
(2007).
3. L. D. Keigwin, S. J. Lehman, M. S. Cook, Eos 87, PP44A07
(2006).
4. K. Hughen et al., Science 303, 202 (2004).
5. J. R. Toggweiler, J. L. Russell, S. R. Carson,
Paleoceanography 21, PA2005 (2006).
6. B. B. Stephens, R. F. Keeling, Nature 404, 171 (2000).
7. R. F. Keeling, B. B. Stephens, Paleoceanography 16, 112
330 (2001).
8. J. F. Adkins, K. McIntyre, D. P. Schrag, Science 298, 1769
(2002).
Published online 10 May 2007;
10.1126/science.1142326.
Include this information when citing this paper.
An advance in DNA sequencing is a crucial
component of a rapid, precise, and relatively
inexpensive way to identify transcription
factor binding sites at a whole-genome level.
Site-Seeing by Sequencing
Downloaded from www.sciencemag.org on June 7, 2007
PERSPECTIVES
Stanley Fields
very few years, a new technology comes
along that dramatically changes how
fundamental questions in biology are
addressed. The impact of the technology is not
always appreciated at first—when it is used
only by those involved in its development—but
becomes clear once the technology begins to
spread to the broader scientific community. A
well-known example is the DNA microarray,
which became widely available to biologists
about a decade ago and has since been applied
to an ever-expanding set of questions such as
E
The author is in the Departments of Genome Sciences and
Medicine, Howard Hughes Medical Institute, University of
Washington, Seattle, WA 98195–5065, USA. E-mail:
[email protected]
determining the profile of genes expressed in a
specific cell type. Now it is ultrahigh-throughput DNA sequencing that is making the transition from development to widespread use.
Johnson and colleagues are in the vanguard of
this movement. On page 1497 of this issue (1),
they report that an advanced DNA sequencing
technology (from Solexa/Illumina) can be used
to identify all the locations in the human
genome where a specific protein binds. They
do this with a speed and precision that goes
beyond what has been achieved with previous
technologies.
DNA-binding proteins control transcription, replication, DNA repair, and chromosome segregation. Given the importance of
these proteins, identifying their binding sites
www.sciencemag.org
SCIENCE
VOL 316
Published by AAAS
throughout the genome has occupied much
attention in recent years. The most common
method of locating these sites within a living
cell is known as chromatin immunoprecipitation (ChIP). In this approach, cells are treated
with a reagent, typically formaldehyde, that
crosslinks protein and DNA, and then the cells
are lysed. Chromatin (the complex of proteins
and DNA in chromosomes) is isolated, the
DNA is sheared into small fragments, and an
antibody is added to precipitate the protein
and its associated DNA. The DNA that is
liberated after reversal of the protein-DNA
crosslinks is then analyzed. In the initial uses
of this method, researchers analyzed the DNA
to determine whether single genes were
enriched by the immunoprecipitation.
8 JUNE 2007
1441